THE SEARCH FOR COLLECTIVE VARIABLES IN ANIMAL BEHAVIOR

I. Golani

Department of Zoology, Tel Aviv University, Ramat Aviv, Tel Aviv, Israel

In my talk I will make the following points:

  1. In the study of behavior, there is a necessity for a systematic justification of what is to be represented and what is to be measured.
  2. A good description is economic , informative, predictive, and therefore also of explanatory power.
  3. To obtain a good description one should be able to choose and shift deliberately among a variety of coordinate systems, between discrete and continuous representations, and between symbolic and rate-dependent representation.
  4. Once a good descriptive model is obtained, it sensitizes the observer to similarities between seemingly very different behaviors, as well as to differences between "similar" behaviors.
These points will be illustrated by examples from interactional behavior (ritualized fighting in honey badgers), normal locomotor behavior (infant rats) and drug-induced locomotor behavior (3 dopamine stimulants). Video tapes of these behaviors will be accompanied by discrete symbolic representations of the movements of the parts of the body (Eshkol Wachmann Movement Notation analysis), and continuous phase space representations based on automatic tracking. Both methods reveal collective behavioral variables which in turn disclose specific sensorimotor algorithms.

Paper presented at Measuring Behavior '96, International Workshop on Methods and Techniques in Behavioral Research, 16-18 October 1996, Utrecht, The Netherlands